Stories from HQ

The Shipping Fees Saga

[for context: I work at Skroutz, an e-commerce marketplace]

It was early February; I was part of a workgroup trying to figure out what hindered our growth. Someone pointed out our shipping fees.?

They said “No one thinks rationally about shipping fees. We detest them so much that we are willing to drop the purchase altogether if we deem the shipping fees too high. Think of users who buy from us once or twice a year and are not members of Skroutz Plus. What kind of impression do they get today? We shouldn’t be too expensive, even for the infrequent user.”

We had already run experiments showing us how important covering the shipping costs was. Our most recent example was Sneakers (the shoes). We ran a month-long promo “free shipping for products >70€” and overturned a 6-month flattish category performance to a stunning +24% YoY.?

You might be wondering - Why did you set your fees so high in the first place? Why didn’t you run more promos?

Fair questions. We didn’t consciously end up with overpriced fees, it happened slowly, over time, like a boiling frog, improving one cost metric after the other, ending up with a scheme that had minimum investment and unfortunately minimum growth. According to the Code of Conduct for the Greek Trade (and common sense), if an offer is always on, it’s not a promo it’s a policy.

And that was the moment when we decided to silently change our policy and recalculate the core shipping fees of 12 million SKUs. As fast as we could.?

The process was so extraordinary that it’s worth a new post itself but I’ll try to paint a quick picture here:?

• Research what “average market” shipping fees are for each of the ~4K categories. Essentially what the user expects to pay in the same fashion as Jakob’s Law of UX but in delivery fees.?
• Assign all inventory products into predefined weight and price buckets. Get and review as many internal data points as a Google sheet can handle for each bucket.?
• Design a policy based on user and merchant location. Make it approach courier companies' price lists. Set it as the base default policy and use it as a rule of thumb for the next step.
• Adjust each bucket's base policy regarding shipping fee value and free shipping thresholds. Align the policies across similar product types.
• Run several simulations, including potential behavior changes, and see how much it will cost
• Develop the infrastructure, assign the new fees
• Monitor the costs, ensure no catastrophes?
• And finally, after a couple of months, assess the results and optimize

Many organizations would set it up as an official internal project. They’d assign it to a business owner, define the scope and the deliverables, allocate resources, and spend an estimated 6-9 months to run the research, align internally with all stakeholders, and implement the suggested changes. Many meetings and documentation would take place and internal approvals would be necessary along each step. The writers of the Simple Sabotage Field Manual would be thrilled with these procedures*.???

I’m so happy I work in a different kind of organization - one with a Geek culture**. What happened was that, since this was deemed as a super high priority, a task force of 15 skroutzers dropped all other assignments and pledged to solve this problem asap. We created templates, agreed on some tenets to work on, manually assigned 10K shipping fee values, challenged each other in a workshop, and delivered the new suggested prices within 10 days. Operations ran several simulations, engineering created new infrastructure, and with a few back-and-forths, the improved pricing was released within one month from the kickoff.?

If you wonder if it worked - it did. Users appreciated the new policy and trusted us with more purchases than before. And we saved a ton of not-wasted money and time with our commando approach. Mind you, we still kept the loyalty thresholds much lower - for our most frequent users to continue getting the extra benefits.?

And what would happen if it didn’t? Our value “better now than perfect” takes care of it. We would continue to iterate and improve it until we got it right. As long as we are alert and ready to fix anything that goes wrong, we are not remotely afraid of making mistakes. Whatever it is, we’ll fix it.?

*Americans published this manual during WW II aimed at people living in Norway and France to help them win the war. The main idea was: “If you want to help win the war, inflict standard corporate operating procedure on the other side”. This is also probably worth a separate post.?

**As defined in the book “The Geek way” by Andrew McAfee . I highly recommend it.